This is the preprint version.

Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

Nuno Venturinha - Epistemic Contextualism, Subject-Sensitive Invariantism and Insensitive Invariantism:

Anything Else?

This paper aims to provide an overall picture of what lies at the heart of my book Description
of Situations: An Essay in Contextualist Epistemology (2018) although the examination ranges even
further afield. Section 1 contrasts epistemic contextualism with what I call context-sensitive
objectivism (CSO). Section 2 discusses a major strand of thought in contemporary epistemology
that also reacts against traditional contextualism: subject-sensitive invariantism. Finally, section 3
explores insensitive invariantism, including Williamson’s anti-sceptical version of the latter. I argue
that CSO has considerable advantages over each of them.

1. EPISTEMIC CONTEXTUALISM

Epistemic Contextualism (henceforth EC) is standardly taken to be concerned with

knowledge attributions and to have as its main feature the indexicality of “know”. Traditional EC
theorists seek to explain this predicate in terms of the same semantic factors that account for our
use of gradable adjectives which, given their vagueness, can be modified by intensifiers and
downtoners. For example, to talk about a cycle route as being “long” is evidently vague. If you are
just an occasional cyclist, 50 miles may look like a very long ride, but someone well-trained could
say instead that it is a fairly long distance. Similarly, the idea is to regard sentences of the form “S
knows that p” as obeying to different epistemic standards that are made salient in the
conversational context. These standards can be higher or lower, that is, more or less epistemically
demanding, and will determine in each case what counts as “knowledge”—a totally relative
concept from the contextualist point of view. According to Stewart Cohen, the earliest major
proponent of EC,1 “one speaker may attribute knowledge to a subject while another speaker denies
knowledge to that same subject, without contradiction” (1987, 3; 1988, 97). Keith DeRose, another
leading EC theorist, considers that “[t]his lack of contradiction is the key to the sense in which the
knowledge attributor and the knowledge denier mean something different by ‘know’” (1992, 920).
But to what extent can it be legitimately argued that there is no contradiction? In Description of
Situations (2018, ch. 1) I claim that the contradiction is unavoidable if a strong conception of
knowledge is assumed. EC, of course, is directed exactly against such a conception putting virtually

1 David Lewis’ 1979 “Scorekeeping in a Language Game” was undoubtedly a forerunner of EC but it would take him
several years to offer a comprehensive account as found in his 1996 “Elusive Knowledge” (see Lewis 1983 and 1999).

1
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

all the emphasis on the rules that govern our practical reasoning. As Patrick Rysiew emphasizes,
“in itself, EC is silent about knowledge” (2011, 111, fn. 1). However, if EC is also committed, as its
proponents forcefully defend, to eradicating radical scepticism,2 then critics like Richard Feldman
(1999, 2001, 2004) and Crispin Wright (2005) are right in stressing that a relativist account of
“know” leaves the sceptical challenge untouched. The argument that radical scepticism will only
arise in high-standards contexts but not in low-standards, ordinary ones looks much more like an
inadmissible capitulation to the sceptic, who is perfectly happy to make demands that are
exceedingly difficult to meet, than like a proper solution to the problem.
Let me illustrate some of these worries with an example. Ralph has been a rock fisherman
his whole life and he is used to observing the phases of the moon to predict high and low tides.
For someone as experienced as him, tide charts are almost unnecessary. Ralph knows that spring
tides occur during the full and new moons, and he always gets his best catches when the tide is
rising. But is it licit to say that he really knows it? Ralph has only an elementary education and
cannot give a full account of the gravitational interactions between the earth, the moon and the
sun. His understanding comes from the regularities he has been able to identify from experience.
In comparison with an astronomer or a geophysicist, who can explain the various forces involved
and provide detailed calculations for their conclusions, Ralph’s knowledge suddenly becomes
ignorance. So should we come up to Ralph and say he does not know what he apparently knew? I
think Ralph would simply reply: “Bring me those scientists and I’ll give them a fishing lesson!”
Ralph’s simplified picture of the intricate phenomena at stake serves a practical purpose and it
completely fulfils this end. It is patently clear, however, that Ralph would be at a loss to explain
long-term variations such as draconic spring tides, which occur roughly every 9.3 years (see Wilson
2012; Wilson and Sidorenkov 2013). Should we then assert, as Plato did nearly twenty-five
centuries ago in the Theaetetus, that knowledge must be accompanied by an explanation? I see no
other way if “knowledge” is understood in this strong sense, but even the scientific explanation
has its limits. There can obviously be much longer-term variations at a macro-temporal level which
involve aspects that are still to be grasped or that will never be grasped.
What is the EC theorists’ take on this? They maintain that it all depends upon the context
in which the attributors find themselves. An attributor whose stake is low and another whose stake
is high will respectively say that Ralph knows and does not know what is happening for the simple
reason that, in DeRose’s words, they “mean something different by ‘know’”. Indeed, it has never
been an issue for Ralph to offer a systematic account of tides. He just wants to ordinarily know his

2 DeRose’s 1995 “Solving the Skeptical Problem” makes a clear and unambiguous statement of such intention (see
DeRose 2017, ch. 1). See also Cohen 1999 and 2000.

2
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

way around them and have the best fishing. An unusually large, draconic tide can catch him every
decade or so but he is also able to notice its long-term consistency and err on the side of caution.
He does not scientifically know why it happens; he only ordinarily knows that those extreme tides
occasionally happen. On the other hand, astronomers and geophysicists understand “know” in a
specialized way. For them, it means to integrate lots of technicalities into a coherent theory, which
is completely alien to Ralph’s mind when he sits on a rock with his fishing rod. But, despite EC’s
ingeniously crafted argument, do not the attributors presuppose that both Ralph and the scientists
are knowing something by means of making approximations to its truth? This way of speaking is
essentially uncongenial to any EC theorist because the basic assumption of contextualism is
precisely, as Kevin Hermberg sensitively put it, “that truth and knowledge are relative to a specific
social context and thus that there is no such thing as objective truth arrived at by cognizers” (2011,
163). Description of Situations challenges this relativist, subjectivist view and proposes a radically
different perspective in epistemology: what I have called elsewhere Context-Sensitive Objectivism,
which I abbreviate as CSO (Venturinha forthcoming a). CSO is not driven by the mere context-
sensitivity of knowledge attributions but rather by the objectively context-sensitive basis of our
“claims about knowledge”—what Rysiew (2011, 111, fn. 1) stresses that EC is not interested in.
This may seem contradictory but I am not alone in thinking that a compatibility between context-
sensitivity and objectivism is possible. For example, in reflecting upon the “contextually sensitive”
nature of “chance ascriptions”, which are conceived within an “objectivism about chance”, Toby
Handfield avers apropos of context-sensitivity that “although it might mean that what proposition is
asserted by a given sentence may depend, in part, on subjective factors, the truth conditions for the
proposition asserted need not depend on subjective factors” (2012, 123). Let me briefly sketch the
fundamentals of this idea.
CSO assumes that there are n epistemic possibilities (EP) inherent to a state of affairs p.
By the epistemic possibilities of p, I mean the various ways in which p can be objectively known,
regardless of whether p is eventually known in a context C at a time t by a subject S.3 It follows
from this definition of EP that the knowability of p rests on mind-independent facts. In order to

3 This notion of “epistemic possibilities” is totally different from DeRose’s, whose concern lies in “possibilities of the
kind that sentences of the form ‘It is possible that P’ […] typically express” (1991, 581). Recently, Scott Aikin and
Thomas Dabay have suggested in a way similar to DeRose that “[s]omething is an epistemic possibility only if we don’t
know it doesn’t obtain” and formalized this principle in modal terms as follows: “EP: E◊p ⸧ ~K~p” (2019, 119, my
emphasis). In truth, this had already been propounded by Ian Hacking in his initial discussion of the issue when he
wrote: “a state of affairs is possible if it is not known not to obtain, and no practicable investigations would establish
that it does not obtain.” (1967, 149). Handfield (2012, 24, 71) also discusses “epistemic possibility” along these lines.

3
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

preserve the objectivity of p when it is accessed in a given C at some t by a certain S, I see no need
to distinguish, as Husserl did, between a situation and a state of affairs.4 The recognition of what
can be objective in p cannot be done at the expense of drawing an intangible line separating p and,
say, p*, its decontextualized version. What CSO vindicates is that all aspects about p that can be
known by S in period t resulting from C are p-subordinated. Note that, like Roderick Chisholm, I
am using the propositional variable p to denote a state of affairs (see Chisholm 1986, 30 et passim).5
Therefore, within the framework of CSO, the EC-type sentence “S knows that p” depends on
there being a corresponding state of affairs p that is propositionally expressible. Again, this
propositional expressibility of p is psychologically independent from S. This is what allows the
sentence “S knows that p” to be ultimately true or false. What makes it true or false can neither be
the attributor’s “own reasoning ability”—namely when the subject and the attributor are the
same6—nor a consensus within epistemic groups, as suggested by Cohen (1986, 579; 1987, 15) and
other contextualists. “S knows that p” is true iff S’s knowledge matches one or more of the n
epistemic possibilities belonging to p. Error, on the contrary, occurs when there is a mismatch in
relation to p. Context-sensitivity matters only insofar as it captures the fluidity of reality, whose
ever-changing nature excludes immutable truths. It is in this sense that, I am convinced, CSO is
able to avoid what John Turri, in commenting on Michael Williams’ rejection of “epistemological
realism”, describes as “[t]he realist’s major mistake”, which “is to suppose that every belief has an
utterly unchangeable epistemological character” (2014, 29). Turri goes on to say that “in everyday
life, the evidential requirements for beliefs ‘shift with context’” concluding that “[t]here are no
fixed and immutable relations of epistemic priority” (ibid.). The anti-realist’s major mistake, as I
see it, is to suppose that context shift cannot have an objective basis. This was what led Williams
to reject that we should believe “that a proposition has a definite epistemological status simply in

4 Husserl’s distinction between “situation” (Sachlage) and “state of affairs” (Sachverhalt) is explained in his rather
idiosyncratic style as follows: “What we call a situation […] appears here merely as the passively preconstituted foundation,
qualitative or relational, of all these states of affairs; but subsequently, if the states of affairs have been constituted and
objectified in an original predication, this foundation can be apprehended objectively as the identical situation which
underlies them.” (1973, 241) Wittgenstein also uses these two terms in the Tractatus, but he takes a more monist
approach seeing “a situation in logical space” as “the existence and non-existence of states of affairs” (1961, 2.11). I
explore this issue further in chapter 2 of Description of Situations.
5 Alluding to “Frege’s use of ‘thought’ or ‘Gedanke’”, Chisholm argues that “propositions” can be regarded as “a
subspecies of states of affairs” (1986, 29). Frege actually occupies a central place in chapters 4 and 5 of Description of
Situations.
6 This relation is expressed in sentences of the form “I know that p” and, as I discuss below, its subjectivity was
explored by some theorists as a different kind of contextualism.

4
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

virtue of being ‘about the world’” (1988, 425).7 It is little surprise that Williams (2018, 2019a,
2019b) puts too small a price on Wittgenstein’s so-called “hinge propositions” of On Certainty,
which in Description of Situations I take to be the best candidates for those propositions “about the
world”.8 But let me resume my argument.
If CSO is right, then Ralph will know a tidal occurrence p if he can estimate at least n – 1
epistemic possibilities of p, but he will not be able to offer a comprehensive estimation of p as
astronomers and geophysicists will since the latter depends on an articulation in a justified theory
of the epistemic possibilities n – 1, n – 2, n – 3, …, ideally reaching n – n. I can hear the EC theorist
grumbling: “So what’s new if they all know? That’s the point!” However, Ralph does not know p
because he himself or someone else attributes this knowledge to him on the presumption that to
“know” means in this context to estimate fortnightly spring tides. He does know p because he
knows something about p that originates from p. When compared to a scientist’s knowledge,
Ralph’s knowledge looks extremely narrow. But the most important thing is that they do not
“mean something different by ‘know’” for on their different planes they are making
approximations to the truth of p. If knowledge were a concept determined by the epistemic
standards of an individual or a social group, and not by the states of affairs themselves, we would
have to accept unreasonable claims to know made by demented persons or conspiracy theorists
like flat earthers. As the radical sceptic has no qualms about raising doubts about what normal
people take for granted, defenders of a flat earth make use of the most perverse arguments to call
into question all scientific evidence that the earth is spherical. And, astonishing as it may seem, all
that EC can say about flat earthers and radical sceptics is that, given their unusually high epistemic
standards, they may well be right. We would be tempted to agree with Timothy Williamson that
“contextualists are apt to console themselves with the thought that although most denials of
‘knowledge’ in [a] context of scepticism are correct, in everyday contexts many assertions of
‘knowledge’ are also correct” (2005a, 689). However, what Williamson acutely observes is “the
gravity of the situation in which the sceptic has put contextualists, on their own analysis” (ibid.).
Let me adapt an example Williamson gives to illustrate this point—there being a whiteboard in

7 What Williams characterizes as the sins of epistemological realism is exactly what I take to be its virtues, namely:
“To treat ‘our knowledge of the world’ as a genuine totality, as even a possible object of wholesale assessment, is to
suppose that there are objective epistemological relations underlying the shifting contexts and standards of everyday
justification. It is to assume that context-sensitivity does not go all the way down, there being rather an underlying
objective structure of justificational relations that philosophical reflection brings into view and which allows us to
determine, in some general, uniform way, whether we are entitled to claim knowledge of the world.” (1988, 425)
8 I respond to Williams 2019b—which is a commentary on Description of Situations—in Venturinha 2019.

5
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

the room within the context of an epistemology seminar and outdoors—to the case I was
discussing and follow the direction of his line of reasoning. If someone affirms, against flat
earthers, “Scientists know that the earth is round”, the factivity of knowledge implies that the
analytically entailed sentence “The earth is round” be true in both the flat earthers’ world (even if
they take this sentence to be false) and the scientists’ world.9 Either the earth is flat or it is round!
But a scrupulous contextualist, in the face of flat earthers’ criticism, will dodge the choice and
contradictorily assert “The earth is round and I don’t know that the earth is round”. Mylan Engel
Jr. summarizes well the conundrum in which EC is trapped when he defines “the metalinguistic
turn”:

First, contextualists maintain that there is no correct context-independent standard of

knowledge. Consequently, there is no context-independent fact of the matter as to whether
or not S knows that p. Since there is no fact of the matter as to whether or not S knows
that p outside a context of ascription, they maintain that we epistemologists should drop
all talk about whether or not S knows that p. Our focus, instead, should be on whether
sentences of the form ‘S knows that p’ are true in some specified context of ascription.
(2004, 207)

EC thus needs reasonable standards, but who can set them except those who make first-
person knowledge claims or attribute knowledge to others always within an epistemic group—
even if its epistemic merits have never been questioned? A perfect world would certainly be one
governed by what DeRose has called “pure reasonableness”, but deep disagreements show that
we are still far from such a stage of civilizational evolution.10 DeRose admits that in a certain
conversation

the matter of which of our speakers is speaking the truth depends on facts about what are
the reasonable standards for them to use in their situation. And, of course, this opens up
a whole host of questions that I won’t even begin to address about what makes standards
the reasonable ones to use. (2009, 142).

He writes a bit further on:

9 On the factive character of knowledge, see also Williamson 2000, ch. 1, and, specifically directed against
contextualism, 2001, 26 ff.
10 DeRose’s attempt to answer Williamson’s concerns discussed above evokes precisely the “reasonableness view”
(2017, 129).

6
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

It’s good for speakers to use reasonable standards (or, more generally, reasonable scores),
of course. But if they opt for unreasonable standards, I’m inclined to think the truth-
conditions of their claims then reflect those unreasonable standards that they are
indicating. (Ibid.)

I am convinced that this brief survey suffices to realize that, all in all, EC creates more
problems than it solves. This is unfortunate because context matters to a great extent, only not
perhaps in the way in which EC theorists would have wished. Criticisms notwithstanding, EC still
attracts many devoted followers who, unwilling to pay the price of realism, happily subscribe to a
view that, in Paul K. Moser’s evasive words, “finds the basis of epistemic justification in a social
consensus of some sort” (1989, 183, my emphasis). But I do not think that epistemologists should
be satisfied with such a muddle about what is to know, especially if they are committed to solving
the sceptical problem instead of sweeping it under the rug.

2. SUBJECT-SENSITIVE INVARIANTISM

One alternative to EC that is also (at least to a certain extent) contextualist and that I did
not explore in Description of Situations is so-called Subject-Sensitive Invariantism (SSI). The label
comes from DeRose (2004; 2009, ch. 6), who criticizes the views of authors such as Jeremy Fantl
and Matthew McGrath, the first to advocate against contextualism that “a subject’s pragmatic
situation may affect her justification” (2002, 70), John Hawthorne, with his “(subject-)sensitive
moderate invariantism” (2004, ch. 4), and Jason Stanley, who speaks more specifically of “interest-
relative invariantism” (2005, chs. 5-7). What is from the outset rejected by SSI, as Hawthorne
explains, is “ascriber-dependence”, which “forces the thesis of context-dependence” (2004, 157).
More specifically, Hawthorne writes,

it forces the conclusion that two ascribers may be looking at a single subject at the same
time and one truly say[s] ‘He knows that p’, another ‘He doesn’t know that p’.
Contradiction is avoided by claiming that the verb ‘know’ expresses different relations in
the mouths of each ascriber. (Ibid.)

The solution, then, is “to consider the deliberative context of the subject” and this means
being sensitive to its “practical environment” wherein rests “the truth of knowledge ascriptions”

7
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

(ibid., 180). On this view, it would not be possible to have, as in EC, someone ascribing knowledge
to Ralph on the basis of low epistemic standards (e.g. his fishing mates) and someone else denying
it on the basis of high epistemic standards (e.g. a scientist listening to a conversation between
Ralph and his fishing mates) without contradiction. What is relevant are only, in Stanley’s jargon,
“the subject’s practical interests” (2005, 122). Thus, if Ralph is merely interested in predicting the
occurrence of spring tides every fortnight and he succeeds in doing so, then he knows it. The same
holds true for the astronomers and geophysicists whose interest lies in achieving scientific
understanding of tidal phenomena. As in EC, in SSI everyone knows in the last analysis. But
whereas EC can trust that the knowledge attributor will appeal to some “reasonable standards”,
SSI does not enjoy those standards inasmuch as the attributor’s context is considered epistemically
irrelevant. Elke Brendel offers a useful snapshot of what is at stake when she avows that “at any
given time t, a knowledge ascription ‘S knows that p’ has a fixed truth value due to the fixed
practical interests of S at t” and this means that “‘know’ is a univocal knowledge relation in SSI”
(2012, 34). Accordingly, there is no variability of “know” between contexts of attribution—and
hence the invariantism that is appended to the subject-sensitive nature of our knowledge claims.
Taking into account, as Jonathan Schaffer puts it, that SSI theorists “claim to capture the
contextualist data without the shifty semantics” (2006, 87), then it could seem that an invariantist
view might help my case and that SSI would be closer to CSO than EC is. But to lay all the
emphasis on the subject is too dangerous a step for epistemologists to take. If knowledge is
construed in terms of our “needs and interests”, as insisted on by Fantl and McGrath (2002, 71),
one of the most vexing outcomes of SSI will be the relegation of specialized knowledge to just
another point of view. Here is how Schaffer lucidly approaches this question:

In general, the social role of the expert is to serve as a reservoir of knowledge. This requires
a stability in one’s pool of knowledge that is not compatible with SSI. The social status of
expertise cannot fluctuate as the stakes rise and fall. For instance, one cannot gain in
expertise by suddenly not caring about the topic. So I conclude that the status of expertise
is not sensitive to what is at stake for the subject. (2006, 97)

At the core of SSI lies an idea of ready-made knowledge as what fuels all of our practices.
There is definitely something epistemically important in the way we know how to do the most
varied things. However, that I know how to walk does not entail that I know that which
biomechanically allows me to walk. It is no coincidence that Stanley, together with Williamson,
has questioned the Rylean view according to which “there is a fundamental distinction between
knowledge-how and knowledge-that”, viewing the former as “simply a species” of the latter

8
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

(Stanley and Williamson 2001, 411).11 To investigate the manner in which our propositions hinge
on more basic ones that are not, to speak like Duncan Pritchard, “in the market for knowledge”
(2016, 77 et passim) may be, so Description of Situations suggests, a promising path. But the kind of
Moorean certainties that we know without knowing supplies no “interest-relative” standard
whatsoever. Looking at the matter from the perspective of EC, seen as the only contender “able
to provide a socially fitting conception of knowledge”, Schaffer strongly argues that “knowledge
must not be sensitive to what is at stake for the subject, but must rather be sensitive to what is in
question for the attributor” (Schaffer 2006, 100).12 CSO, in turn, is in a very particular sense
context-sensitive, but it rejects that either epistemic variantism across contexts or subject-
sensitivity within a specific context can contribute to an inclusive view of knowledge which must
go beyond the social sphere.13

3. INSENSITIVE INVARIANTISM

What about Insensitive Invariantism (II for short)? Can it do better than EC and SSI?
There are, no doubt, some points in favour of II. Above all, it is not a relativistic view amenable
to attributor contexts or subject contexts. Quite the contrary, it is an epistemological view which
assumes that this thing we call truth is not laid down by ourselves and that knowledge involves
apprehending something about reality. In this sense, it looks very similar to CSO, at least to its
objectivistic component. In fact, compare my characterization of CSO to how Wayne A. Davis
justifies his preference for II over both EC and SSI:

I advocate classical or insensitive invariantism, on which the truth conditions of ‘S knows p’

do not vary with truth-independent factors.
[…] There is a wide variety of uncontroversially context-sensitive terms, but all behave
differently from ‘knows p’ in several ways. For example, ‘is flat’ and ‘is heavy’ allow
comparisons and relativizations, as do ‘knows Paris’, ‘knows how to dance’, and ‘is justified
in believing p’. But ‘knows that p’ cannot be qualified by more or better, and we never say

11 On this topic, see in addition Stanley 2011. Williamson, however, does not present himself as an SSI-er, proposing
in place of EC or SSI an “insensitive invariantism”, described as “the view on which the epistemic terms at issue
undergo no shift in the standard for their correct application” (2005b, 213). See more below.
12 Other relevant criticisms of SSI can be found in Blaauw 2008, Brown 2008, Blome-Tillmann 2009 and Baumann
2016.
13 The subtitle of my book, An Essay in Contextualist Epistemology, must thus be taken with a pinch of salt.

9
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

things like: He knows it’s true by (or relative to) low standards, but not high standards; He knows it’s
true compared to Mary, but not Jane; or He knows it’s true, but not perfectly. (2017, 219)

From Davis’ insensitive invariantist account, what Ralph supposedly knows about spring
tides would only be permissible if it were expressed under the form of the examples given at the
end of this quotation. Ralph, however, does not evidently say that He knows it’s true that spring tides
happen every fortnight by (or relative to) low standards, but not high standards, that He knows it’s true that spring
tides happen every fortnight compared to his fishing mate Stubb, but not scientists or that He knows it’s true that
spring tides happen every fortnight, but not perfectly. The point made by II-ers is precisely that Ralph sees
no further possibilities of knowing that p and unjustifiably takes his understanding of the
phenomenon to constitute knowledge. In other words, he is unaware of his ignorance. To speak about
low standards for knowing that p is already the result of our recognition of there being higher
standards for doing it. When we are in such an epistemic position, the II-er maintains, only high
standards are admissible and there is simply no room for low standards. The problem for II is that
what we now regard as the (high) standard for knowledge can in the future turn out to be merely
a lower standard or even an error, with the (high) standard being replaced by a new one. Just think
about what Ptolemy and his followers, unaware that they were misconceiving the heavenly bodies
as revolving around a stationary earth in circular orbits, considered as high standards in comparison
with Copernicus, who within his heliocentric model nevertheless kept the idea of circular motion,
or Kepler, who showed that planetary orbits are elliptical.
Our position, like Ralph’s and anybody else’s, is always viewed as the standard, even if we
recognize that it can be revisable. In that case, the standard will be to hold a revisionary posture,
one that ultimately leads to an awareness of continuous possible ignorance in the midst of which
knowledge must be permanently treated as provisional. Does this imply adopting a form of
scepticism? Williamson actually highlights that “[s]ome insensitive invariantists are sceptics”, given
that for them “‘know’ invariably refers to a maximal epistemic standard that we cannot meet”
(2005b, 225). He thus proposes an “anti-sceptical insensitive invariantism” (ASII for short), within
which “‘know’ invariably refers to an epistemic standard that we can and do meet quite easily”
since “everyday ascriptions of knowledge are often true” (ibid.). Williamson’s strategy is to replace
the II-er’s “attempt to explain the illusion of knowledge” with another task proper to the ASII-er,
which consists of explaining “the illusion of ignorance” (ibid.). Williamson is absolutely right in
arguing that there are limits to what can be doubted and that philosophical scepticism is
responsible for introducing a radically unnatural suspicion. His diagnosis that “an illusion of
epistemic danger” is the “result from exposure to lurid stories about brains in vats, evil demons,
painted mules, or gamblers who bet the farm” (ibid., 226) seems to me perfectly judicious.
10
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

However, while I fully subscribe to the anti-scepticism of this view, Williamson’s appeal to
“practical reasoning” (ibid., 227 ff.) makes AS II coincident with SSI.14 What is distinctive of any
insensitive invariantist description is indeed the denial that “practical” reasons can affect the
knowledge relation. Michael Hannon stresses this point when he writes:

Insensitive invariantism is the main source of resistance to contextualism. According to

this view, what counts as being in a sufficiently good epistemic position to know does not
vary with practical facts about the context. Whatever the subject’s or the attributor’s (or
someone else’s) practical interests might be, there is some good epistemic position in which
an agent must stand with respect to a proposition in order for that agent to know it. (2019,
165)

Hannon brings to the fore an inevitable struggle against contextual factors that is typical
of II-ers, for whom the “good epistemic position” is not relative to standards. So what should
distinguish ASII-ers? If Williamson has in mind, for instance, that Ralph cannot doubt that a spring
tide happens every fortnight or (imagining a theoretical scenario or that Ralph suddenly suffers
from some mental disorder) that the fishing rod in his hand and his own hands are real, I would
completely agree. But I do not see how—within Williamson’s framework—uttering those things
can be conceived without the attributor of EC or the subject of SSI. And since any form of
invariantism per definitionem excludes EC’s variantism, ASII can only fit in SSI. As a matter of fact,
Williamson’s KPR principle, which states that “A first-person present-tense ascription of ‘know’
with respect to a proposition is true in a context iff that proposition is an appropriate premise for
practical reasoning in that context” (2005b, 227; see also 231), is manifestly subject-sensitive
invariantist and as such not anti-sceptical at all. Ralph does not know what he is convinced he
knows about tides only because what he knows is an appropriate premise for his practical reasoning.
Definitely it is, but that he cannot even imagine to reason differently in his daily practice would
neither detain the sceptic from raising Humean doubts about Ralph’s inductively inferred beliefs
nor, more importantly, the scientist from claiming another level of knowledge that can be attained.
The closure of Williamson’s ASII—as that of SSI—cannot thus originate more than what Ian Evans
and Nicholas Smith called “a (compelling) fallibilist invariantist explanation of our intuitions”
(2012, 69).
Yet, unlike Evans and Smith, I do not find fallibilism compelling. Following Wittgenstein
(1974) and Lewis (1999), I argue that a fallibilist view, equally defended by contextualists like

14 See note 11 above.

11
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

Cohen (1988), does not provide the best shield against radical scepticism and that only infallibilism
can coherently demonstrate the illogical character of the sceptic’s manoeuvres.15 The advantage of
an infallibilist perspective is that it is capable of being intransigent about what cannot be subject
to doubt without sacrificing revisionism, as the reflex of a natural progress of knowledge.
Pritchard’s interpretation of Wittgenstein’s On Certainty offers an excellent framework in making
room for a distinction between “über hinge commitments”—intimately connected with “anti-
skeptical hinge commitments” and formulated in “über hinge propositions” or “anti-skeptical
hinge propositions”—and “personal hinge commitments”—with their “personal hinge
propositions” (2016, 95-97). While the former “hinge commitments” are stationary, the latter are
susceptible to change as long as they mirror our embedment in a world where culture and science
evolve. On this view, which is at the bottom of CSO, Ralph can perfectly claim that he knows that
spring tides occur every fortnight and at the same time the scientists can claim that they know how
to explain tidal phenomena according to what is scientifically established so far. As pointed out,
this could also be the outcome in EC but, contrary to what EC theorists admit, what is known is
presupposed as being known “about the world”. Without this presupposition, which is anchored
to the über hinges as our most elementary natural ties to the world, EC’s response to scepticism
is completely defective. It could be argued that what Ralph knows about tides would then be
comparable to what, say, a seagull knows about them, though I do not see this as constituting an
objection.16 Wittgenstein understands “certainty” to mean exactly “a form of life”, envisaging it
“as something that lies beyond being justified or unjustified, as it were, as something animal” (1974,
§§ 358-359). It should be noted in addition that the epistemic multiplicity permitted by CSO—
akin to the epistemic variability permitted by EC—is not tolerable by either SSI (including ASII) or
II. Whereas the former can only admit Ralph’s low-standards knowledge claims, the latter can do
no more than only admitting the opposite.
Table 1 gives a perspicuous arrangement of the different options available. It is modelled
on Michael Blome-Tillmann’s scorecards that access the compatibility and incompatibility with

15 I elaborate on this in chapters 6, 7 and 11 of Description of Situations, while chapters 8, 9 and 10 are devoted to
different sceptical attitudes. See also Venturinha 2020, in which I defend a Lewisian “non-sceptical infallibilism”
against the fallibilism articulated by Jessica Brown in her 2018, and Venturinha forthcoming b, in which I respond
from a Wittgensteinian standpoint to Anna Boncompagni’s fallibilist approach set forth in a commentary on my book
(Boncompagni forthcoming).
16 In commenting on Ernest Sosa’s initial discussion of the matter (1988), I consider in chapter 3 of Description of
Situations that attributing knowledge to animals cannot just be seen as “metaphorical” insofar as they certainly know
something “about the world”. This leads me to reject in that chapter, after Bolzano, Frege and Lewis, a
correspondence theory of truth, which is necessarily anthropocentric.

12
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

EC and SSI of “temporal embeddings”, “modal embeddings” and “conjunctive ascriptions” (2009,
316). My table is somewhat less ambitious and aims to access only the compatibility and
incompatibility with EC, SSI, II and CSO of Ralph’s low-standard knowledge as opposed to a
higher standard of knowledge:

Table 1
Ralph’s Knowledge
Low Standards High Standards
EC  
SSI  n/a
II n/a 
CSO  

Here one clearly sees that in SSI or II, as Brendel put it, “‘know’ is a univocal knowledge
relation” and that EC, although it has shortcomings, is nonetheless closer to CSO in treating
knowledge as multifarious. Much work is still needed to make a comprehensive epistemological
theory out of CSO. In particular, it must be flexible enough to accommodate very different kinds
of knowledge attributions or claims. But I hope that my arguments have shown that CSO can
provide a more solid theoretical basis than the alternatives proposed in contemporary
epistemology. At least one thing seems certain: its rival theories do not involve less problems.
Many will say, however, that objectivism is an illusion and that we should conform to the idea that
the world is always a result of our projections. Is it not the corollary of quantum mechanics that,
as theoretical physicist Carlo Rovelli eagerly professes, “we must accept the idea that reality is only
interaction”, that it is “less about objects than about interactive relationships” (2016, 18, 41)? Here
I prefer to side with Einstein’s view, old-fashioned as it may look, for whom “quantum mechanics
is logically consistent but […] it is an incomplete manifestation of an underlying theory in which
an objectively real description is possible” (Pais 2005, 455) or, more simply, “a self-consistent but
incomplete description of the objective processes” (Lehner 2014, 331). The fact that general
relativity continues to subsist alongside quantum theory, despite being incompatible with one
another, shows that there is still room for epistemological realism. Those who do not feel attracted
to such a quest for objectivity can perhaps take comfort in the moral subjectivism articulated in
the last chapter of Description of Situations, which is in stark contrast to the previous chapters.

13
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

REFERENCES

AIKIN, S. and DABAY, T. (2019). Pragmatist Anti-Skepticism: at What Price?, in K. McCain and T. Poston
(eds.), The Mystery of Skepticism: New Explorations. Leiden: Brill, 105-127.
BAUMANN, P. (2016). Knowledge across Contexts: A Problem for Subject-Sensitive Invariantism. Dialogue:
Canadian Philosophical Review 55(2), 363-380.
BLAAUW, M. (2008). Subject Sensitive Invariantism: In memoriam. The Philosophical Quarterly 58(231), 318-
325.
BLOME-TILLMANN, M. (2009). Contextualism, Subject-Sensitive Invariantism, and the Interaction of
‘Knowledge’-Ascriptions with Modal and Temporal Operators. Philosophy and Phenomenological Research
79(2), 315-331.
BONCOMPAGNI, A. (forthcoming). On Contexts, Hinges, and Impossible Mistakes. Logos & Episteme.
BRENDEL, E. (2012). Knowledge, Contextualism, and Moorean Paradox, in C. Jäger and W. Löffler (eds.),
Epistemology: Contexts, Values, Disagreements. Frankfurt: Ontos, 15-40.
BROWN, J. (2008). Subject-Sensitive Invariantism and the Knowledge Norm for Practical Reasoning. Noûs
42(2), 167-189.
______ (2018). Fallibilism: Evidence and Knowledge. Oxford: Oxford University Press.
CHISHOLM, R. M. (1986). Self-Profile, in R. J. Bogdan (ed.), Roderick M. Chisholm. Dordrecht: Springer, 3-
77.
COHEN, S. (1986). Knowledge and Context. The Journal of Philosophy 83, 574-583.
______ (1987). Knowledge, Context, and Social Standards. Synthese 73, 3-26.
______ (1988). How to be a Fallibilist. Philosophical Perspectives 2, 91-123.
______ (1999). Contextualism, Skepticism, and the Structure of Reasons. Philosophical Perspectives 13, 57-89.
______ (2000). Contextualism and Skepticism. Philosophical Issues 10, 94-107.
DAVIS, W. A. (2017). Loose Use and Belief Variation, in J. J. Ichikawa (ed.), The Routledge Handbook of
Epistemic Contextualism. Abingdon: Routledge, 218-229.
DEROSE, K. (1991). Epistemic Possibilities. The Philosophical Review 100(4), 581-605.
______ (1992). Contextualism and Knowledge Attributions. Philosophy and Phenomenological Research 52(4),
913-929.
______ (2004). The Problem with Subject-Sensitive Invariantism. Philosophy and Phenomenological Research
68(2), 346-350.
______ (2009). The Case for Contextualism: Knowledge, Skepticism, and Context, Vol. 1. Oxford: Clarendon Press.
______ (2017). The Appearance of Ignorance: Knowledge, Skepticism, and Context, Vol. 2. Oxford: Oxford
University Press.
ENGEL JR., M. (2004). What’s Wrong with Contextualism, and a Noncontextualist Resolution of the
Skeptical Paradox. Erkenntnis 61, 203-231.
EVANS, I. and SMITH, N. D. (2012). Knowledge. Cambridge: Polity Press.

14
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

FANTL, J. and MCGRATH, M. (2002). Evidence, Pragmatics, and Justification. The Philosophical Review, 111(1),
67-94.
FELDMAN, R. (1999). Contextualism and Skepticism. Philosophical Perspectives 13, 91-114.
______ (2001). Skeptical Problems, Contextualist Solutions. Philosophical Studies 103, 61-85
______ (2004). Freedom and Contextualism, in J. K. Campbell, M. O’Rourke and D. Shier (eds.), Freedom
and Determinism. Cambridge, MA: MIT Press, 255-276.
HACKING, I. (1967). Possibility. The Philosophical Review 76(2), 143-168.
HANDFIELD, T. (2012). A Philosophical Guide to Chance. Cambridge: Cambridge University Press.
HANNON, M. (2019). What’s the Point of Knowledge? A Function-First Epistemology. New York: Oxford
University Press.
HAWTHORNE, J. (2004). Knowledge and Lotteries. Oxford: Clarendon Press.
HERMBERG, K. (2011). Husserl’s (even-more) Social Epistemology, in P. Vandevelde and K. Hermberg
(eds.), Variations on Truth: Approaches in Contemporary Phenomenology. London: Continuum, 161-176.
HUSSERL, E. (1973). Experience and Judgment: Investigations in a Genealogy of Logic, ed. L. Landgrebe, tr. J. S.
Churchill and K. Ameriks. London: Routledge and Kegan Paul.
LEHNER, C. (2014). Einstein’s Realism and His Critique of Quantum Mechanics, in M. Janssen and C.
Lehner (eds.), The Cambridge Companion to Einstein. New York: Cambridge University Press, 306-353.
LEWIS, D. (1983). Scorekeeping in a Language Game, in Philosophical Papers, vol. I. New York: Oxford
University Press, 233-249.
______ (1999). Elusive Knowledge, in Papers in Metaphysics and Epistemology. Cambridge: Cambridge
University Press, 418-445.
MOSER, P. K. (1989). Knowledge and Evidence. Cambridge: Cambridge University Press.
PAIS, A. (2005). “Subtle is the Lord…”: The Science and the Life of Albert Einstein. Oxford: Oxford University
Press.
PRITCHARD, D. (2016). Epistemic Angst: Radical Skepticism and the Groundlessness of our Believing. Princeton:
Princeton University Press.
ROVELLI, C. (2016). Seven Brief Lessons on Physics, tr. S. Carnell and E. Segre. London: Penguin.
RYSIEW, P. (2011). Surveys, Intuitions, Knowledge Attributions: Comments on Keith DeRose’s
“Contextualism, Contrastivism, and X-Phi Surveys”. Philosophical Studies 156, 111-120.
SCHAFFER, J. (2006). The Irrelevance of the Subject: Against Subject-Sensitive Invariantism. Philosophical
Studies 127, 87-107.
SOSA, E. (1988). Knowledge in Context, Skepticism in Doubt: The Virtue of our Faculties. Philosophical
Perspectives 2, 139-155.
STANLEY, J. (2005). Knowledge and Practical Interests. Oxford: Clarendon Press.
______ (2011). Know How. Oxford: Oxford University Press.
______ and WILLIAMSON, T. (2001). Knowing How. The Journal of Philosophy 98(8), 411-444.
TURRI, J. (2014). Epistemology: A Guide. Chichester: Wiley-Blackwell.

15
 This is the preprint version.
Please cite from the published paper in Manuscrito 43, no. 3 (2020): 165-186.

VENTURINHA, N. (2018). Description of Situations. An Essay in Contextualist Epistemology. Cham: Springer.

______ (2019). Replies to Commentators. Philosophia, https://doi.org/10.1007/s11406-019-00153-w.
______ (2020). Non-sceptical Infallibilism. Analysis 80(1), 86-95.
______ (forthcoming a). Context-Sensitive Objectivism: Going Deeper into Description of Situations. Logos
& Episteme.
______ (forthcoming b). Replies to Critics. Logos & Episteme.
WILLIAMS, M. (1988). Epistemological Realism and the Basis of Scepticism. Mind 97(387), 415-439.
______ (2018). Wittgenstein and Skepticism: Illusory Doubts, in D. E. Machuca and B. Reed (eds.),
Skepticism: From Antiquity to the Present. London: Bloomsbury, 481-505.
______ (2019a). Beyond Unnatural Doubts: Lessons from Wittgenstein, in E. Marchesan and D. Zapero
(eds.), Context, Truth, and Objectivity: Essays on Radical Contextualism. New York: Routledge, 59-87.
______ (2019b). The Indispensability of Knowledge. Philosophia, https://doi.org/10.1007/s11406-019-
00139-8.
WILLIAMSON, T. (2000). Knowledge and its Limits. Oxford: Oxford University Press.
______ (2001). Comments on Michael Williams’ “Contextualism, Externalism and Epistemic Standards”.
Philosophical Studies 103, 25-33.
______ (2005a). Knowledge and Scepticism, in F. Dretske and M. Smith (eds.), The Oxford Handbook of
Contemporary Philosophy. Oxford: Oxford University Press, 681-700.
______ (2005b). Contextualism, Subject-Sensitive Invariantism and Knowledge of Knowledge. The
Philosophical Quarterly 55(219), 213-235.
WILSON, I. R. G. (2012). Lunar Tides and the Long-Term Variation of the Peak Latitude Anomaly of the
Summer Sub-Tropical High Pressure Ridge over Eastern Australia. The Open Atmospheric Science Journal
6, 49-60.
______ and SIDORENKOV, N. S. (2013). Long-Term Lunar Atmospheric Tides in the Southern
Hemisphere. The Open Atmospheric Science Journal 7, 51-76.
WITTGENSTEIN, L. (1961). Tractatus Logico-Philosophicus, tr. D. F. Pears and B. F. McGuinness. London:
Routledge and Kegan Paul.
______ (1974). On Certainty, rev. edn, ed. G. E. M. Anscombe and G. H. von Wright, tr. Denis Paul and
G. E. M. Anscombe. Oxford: Blackwell.
WRIGHT, C. (2005). Contextualism and Scepticism: Even-Handedness, Factivity and Surreptitiously Raising
Standards. The Philosophical Quarterly 55(219), 236-262.

16